Levels of DNA and/or hemoglobin pyridyloxobutylation were compared in A/J mice or F344 rats treated with a single dose of [5-3H]4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone ([5-3H]NNK),[5-3H]4-hydroxy-1-(3-pyridyl)-1-butanone ([5-3H]4-HPB) or [5-3H]4-(acetoxymethylnitrosamino)-1-(3-pyridyl)-1-butanone ([5-3H]NNKOAc), a compound that generates the proposed pyridyloxobutylating agent in situ upon esterase hydrolysis. The lung and liver DNA samples isolated from A/J mice treated with the various compounds were subjected to acid hydrolysis and the hydrolysates were analyzed for the presence of [5-3H]4-HPB No detectable levels were found in the lung DNA isolated from [5-3H]4-HPB-treated animals, whereas significant amounts of [5-3H]4-HPB were released from lung and liver DNA isolated from [5-3H]NNKOAc-treated mice. The levels of total binding and [5-3H]4-HPB released from the globin isolated from these animals showed a similar trend. That is, low binding levels were detected in the globin isolated from [5-3H]4-HPB-treated animals and significantly higher levels of binding were detected in the globin from the [5-3H]NNKOAc-and [5-3H]NNK treated animals. Comparable findings were obtained in the rat experiments. These studies clearly demon strate that methyl hydroxylatlon of NNK leads to a species that is capable of reacting covalently with nucleophiles in DNA and protein. Thus, the levels of 4-HPB released from DNA and globin can be attributed to the activation of NNK and not to the direct binding of 4-HPB.